rcast_20140411

48
規模・多種多様なデータを扱う為の 検・機械学習技術 株式会社 Preferred Infrastructure エッジヘビーコンピューティング事業部 野健太 [email protected] / @delta2323_ 2014/04/09 @東京学先端科学技術研究センター

Upload: preferred-infrastructure-preferred-networks

Post on 27-May-2015

14.379 views

Category:

Technology


0 download

TRANSCRIPT

2. Preferred Infrastructure, Inc. (PFI) : 20063 : : 30()26/ : PFI / / / TopCoderICPC 2 3. PFI 3 4. 4 5. 5 6. SedueJubatus Data Source Jubatus Sedue 7. PheWAS 8. DNA, WGS/EGSRNA-seq, ChIP-seq 9. Collection Reporting Analytics Action 9 10. Collection Reporting Analytics Action 10 11. Fluentd Flume Kinesis (Amazon) MachineHuman Hadoop S3 (Amazon) Splunk (Splunk) OpenXC (Ford) Mahout Bazil (PFI) AWS (Amazon) Jubatus (NTT, PFI) SAMOA (Yahoo) Qlikview (QlikTech), Tableau (Tableau Software) Bazil (PFI) N. A. 12. 13. CYP3A4/hERG : (HTS) CYP3A4[1] [1] hERG potassium channels and cardiac arrhythmia, Michael C. Sanguinetti & Martin Tristani-Firouzi, Nature 440, 463-469(23 March 2006) (doi:10.1038/nature04710) Fig. 5 13 14. : GGRNA Google-like full text search engine (DBCLS) NCBIRefSeq 13(Zoo) [1] GGRNA: an ultrafast, transcript-oriented search engine for genes and transcripts, Yuki Naito and Hidemasa Bono, Nucl. Acids Res. (2012) 40(W1):W592-W596 Sedue Nucl.AcidsRes.2012 [1] 14 15. GGRNA/GGGenome [] 2U1CPU 62 3.46GHz/192GB GGGenome RefSeq 61 8.6GB 52.4GB DDBJ 92.0 150.8GB 932.2GB hg19 3.1GB 19.0GB GGRNA RefSeq 61 32.4GB 210.3GB DDBJ() 92.0 559.2GB 3192.8GB [] 16. iPS NCBI 16 17. PheWAS 18. X-WAS URL http://www.plosone.org/article/fetchObject.action?uri=info%3Adoi %2F10.1371%2Fjournal.pone.0072737&representation=PDF http://hmg.oxfordjournals.org/content/early/2013/09/06/ hmg.ddt430.abstract http://www.genomics.cn/en/news/show_news?nid=99231 http://psb.stanford.edu/psb-online/proceedings/psb14/hall.pdf http://www.unboundmedicine.com/medline/citation/19048631/ 19. Phenome-Wide Association Study : PheWAS SNPs1 Reverse GWAS vs 1N 2000VanderbiltJoshua C Denny 2010Nature BiotechnologyTrend EMR/PHR GWAS PheWAS 20. PheWAS P P [1] 4/7 2.810-6 0.011 19 < 0.01 [2] 51/77 63 < 4.610-6 [1] PheWAS: demonstrating the feasibility of a phenome-wide scan to discover gene-disease associations Fig. 1 [2] Systematic comparison of phenome-wide association study of electronic medical record data and genome-wide association study data. Fig. 1 21. PheWAS GWAS Phenotype > 22. 23. 24. 25. II 25 26. 26 27. l l l GPS l l AQIUV l 27 28. 29. 29 30. 1 1 30 31. 32. 32 , Edge-Heavy Data: CPS GICTF 2012, http://www.gictf.jp/doc/20120709GICTF.pdf 33. 33 1000 Petabytes/Year > 200 Petabytes In Edge Devices (Surveillance Cameras and Smartphones in Japan) In Huge Computing Cloud (300,000 nodes, each node has 2TB HDD, redundancy is 3) 34. 34 IoT 35. CiscoGE Cisco : Internet of Everything(IoE) IoE1014 4000 l l 76105% l 2013990 35 - White Paper Embracing the Internet of Everything To Capture Your Share of $14.4 Trillion - Industrial Internet: Pushing the Boundaries of Minds and Machines - The Industrial Internet@Work GE : Industrial Internet Industrial InternetGDP 20100150 CTMRI 40025000 36. PheWAS 37. Copyright 2006-2014 Preferred Infrastructure All Right Reserved. 38. (@delta2323_) https://preferred.jp/career/member/oono/ PFI Jubatus Epigenetic 38 39. 1 40. EMR/PHR, 40 twitter 41. Knuth-Morris-Pratt / Aho- Corasick / Boyer-Moore / q-gram / / Suffix Array Genbank 0 1 2 3 102 150 42. Dimensionality Reduction by Learning an Invariant Mapping Raia Hadsell, Sumit Chopra, Yann LeCun, CVPR, 2006 42 43. 43 44. PheWAS 2 45. GWAS Linkage Disequilibrium : LD) SNPs PLAMP in vitro/in vivo 46. GWAS 1000genome projectHapMap Project DBCLS Common Disease TA Manolio et al. Nature 461, 747-753 (2009) doi: 10.1038/nature08494 HapMap Project HP 47. PheWASGWAS 1. 2. SNPs 3. SNPs JoshuaNLP Ninety-nine percent of the work is not in software engineering or coding ICD9 48. PheWAS Medicare, Medicaid HIPAA Act (EHR) HITECH ActEHR (meaningful use) PheWAS